Kimi K3 Pretrain Capabilities Recalibrated
RyanGreenblatt · x · 2026-07-17
The author shared quick test results for Kimi K3 pretrain and later revised their initial judgment.
Key points:
- Initially, they estimated Kimi K3's pretrain capabilities to be close to, or even slightly higher than, Opus 4.8;
- Upon re-testing, they felt this was too optimistic;
- The new conclusion is that Kimi K3's pretrain is roughly between Opus 4 and Opus 4.5, lagging behind Anthropic by about 8 to 10 months;
- They also caution that such tests are better at measuring general pretrain capabilities and may not fully reflect code data quality.
Overall, this is a measured correction of model capabilities rather than simple benchmark chasing.
Related event: Kimi K3 Triggers a Reassessment of Chinese Frontier AI(94 posts)→
More from Models
- GPT 6.1 Terra vs Claude Sonnet 5.5: speed and value vs raw capability — teortaxesTex · 2026-10-27
- Harmonic Teases: AI Taught to Prove Math, 'Something Harder' Coming Soon — satnam6502 · 2026-10-27
- OpenAI quietly swaps Plus/Pro default to cheaper GPT-6.1 Terra and cuts Codex allowance — kimmonismus · 2026-10-27
- 'Never been this uninterested after an OpenAI model release' — theteknosaur · 2026-10-27
- Dev questions reports of Chinese models writing insecure code, citing thin methods and conflicts — dbreunig · 2026-10-27
- Aidan Gomez: LLM probing outside servers was out-of-bounds, not proof of intent — AccBalanced · 2026-10-27